Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (17)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Bomont, P.
Right arrow Articles by Koenig, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bomont, P.
Right arrow Articles by Koenig, M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Human Molecular Genetics, 2003, Vol. 12, No. 8 813-822
DOI: 10.1093/hmg/ddg092
© 2003 Oxford University Press

Intermediate filament aggregation in fibroblasts of giant axonal neuropathy patients is aggravated in non dividing cells and by microtubule destabilization

Pascale Bomont and Michel Koenig*

Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP, 1 rue Laurent Fries, BP 10142 67404 Illkirch Cedex, France

Received November 11, 2002; Accepted February 7, 2003


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Giant axonal neuropathy (GAN) is a severe neurodegenerative disorder characterized by the accumulation of neurofilaments (NFs) in distended axons. GAN corresponds to a disorder of the cytoplasmic intermediate filaments (IFs), since an abnormal aggregation of different IFs has been reported in several cell types, including NFs in neurons and vimentin in fibroblasts. The recent identification of the defective protein, named gigaxonin, now renders possible investigations on the mechanisms that trigger the destabilization of IFs. Although gigaxonin domain organization suggests multiple protein–protein interactions, via the BTB and the Kelch domains, the low amino acid identity with other members of the BTB/Kelch subfamily did not allow hypothesis about its function. In the present work, we studied GAN primary fibroblasts, and show that vimentin aggregation suffers great variation on prolonged culture at confluence and in low serum condition. While neither the microfilament (MF) nor the microtubule (MT) networks are perturbed by vimentin destabilization, we found that the aggregates are in close proximity to the microtubule organizing centers (MTOCs). Moreover, we show that MTs depolymerization induces a total vimentin aggregation in GAN fibroblasts. The results, together with the recent finding of an interaction between gigaxonin and MAP1B, a MT associated protein, suggests that gigaxonin plays an important role in the crosstalk between the IF and MT networks. We found that, when overexpressed, gigaxonin is localized in the cytoplasm but does not colocalize with any of the cytoskeletal networks, suggesting that the presence of the binding partner is rate limiting for proper localization of gigaxonin.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Giant axonal neuropathy (GAN; MIM 256850) is an early-onset and a severe neurodegenerative disorder that affects both the peripheral nerves and the central nervous system. The most frequent form is severe with an age of onset that varies from 18 months to 5 years, leads to a loss of ambulation around adolescence and death of patients around the third decade. One milder form has been reported in a Tunisian (1) and an Algerian family (2), with an age of onset delayed to 6–10 years, a slower progression of the symptoms that allowed clinical examination of a 53-year-old patient.

The defective protein in GAN seems to play a central role in the maintenance of the intermediate filament (IF) integrity. Indeed, ultrastructural studies done on both biopsies and autopsies of GAN patients revealed an abnormal accumulation of the cytoplasmic IFs in various cell types (3–7), implicating four classes of the cytoplasmic IFs (8): keratin (class I and II IFs), vimentin, desmin, glial fibrillary acidic protein (class III) and neurofilament (NF; class IV). Nevertheless, the impairment of the NF integrity, seen both in the central and peripheral nervous systems, seems to be predominant in the pathogenesis of GAN. Biopsy of the peripheral nerves and autopsy of the central nervous system revealed the presence of enlarged axons that present unusual thin myelin sheets according to their size (6,9–12). Such giant axons are filled with abnormally densely packed NFs that lack their normal parallel orientation along the axis of the axone (6,9–12). This specific feature represents useful criteria for the diagnosis of GAN.

Following the study of primary fibroblasts derived from skin biopsies of patients that present vimentin aggregation, it has been proposed that the primary defect implicates the organization of the IFs, rather than resulting in quantitative or qualitative defects of the IFs themselves. This is supported by several studies showing no defect in the rate of synthesis of vimentin, either in two-dimensional gel profiles or in peptide mapping or phosphorylation rate, that determine the state of polymerization of the IFs (13–15).

We recently identified a novel protein, we named gigaxonin as the defective protein in GAN (GenBank accession no. AAG35311) (16). Although the mechanisms by which gigaxonin is implicated in the intermediate filament organization are not known, its domain organization suggests that this is mediated by multiple protein interactions. Indeed, gigaxonin contains BTB and Kelch repeat domains, which are both implicated in protein–protein interactions in the context of other proteins (17–19). Gigaxonin nevertheless shares only 22–27% amino acid identity with other BTB/Kelch proteins (16).

Here, we used cultured primary fibroblasts of patients as a cellular model for GAN in order to characterize and quantify the vimentin destabilization. We show that (i) a minority of the cells presents dense bundles in normal serum condition, (ii) prolonged cultures in both low serum condition and at confluence lead to a great increase of the phenotype and appearance of new forms of aggregates, and (iii) vimentin aggregation, although having no effect on the normal distribution of both the microfilament (MF) and the microtubule (MT) networks, seems to be spatially related to the MT organizing center (MTOC), and to be strikingly aggravated when MTs are depolymerized.

We also investigated the subcellular localization of gigaxonin, by designing specific antibodies against several parts of the protein. Using several cell types, we were unable to detect endogenous gigaxonin on western blots and by immunofluorescence, a result in agreement with the low level of endogenous transcript (16). We found that, when overexpressed, gigaxonin has a cytoplasmic granular pattern that does not colocalize with any of the cytoskeletal networks.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
We studied the IF organization in primary fibroblasts derived from skin biopsies of seven patients suffering from GAN, for whom we identified the corresponding mutations in the GAN gene (see Materials and Methods).

We observed variability in cell morphology and cell doubling between cells but noticed no significant differences between the control and GAN fibroblasts, as well as between GAN fibroblasts. A difference between wild-type and mutant cells was visualized by immunocytochemistry, by the presence of aggregates of vimentin only in some fibroblasts of GAN patients.

Presence of perinuclear ovoid aggregates of vimentin in GAN fibroblasts
In normal serum condition (10% FCS), the vimentin network is filamentous and homogeneously distributed throughout the cytoplasm of control cells (Fig. 1A). With the exception of fibroblasts of patient no. 4, who has a very mild form of the disease, we found that the vimentin network was disorganized in GAN fibroblasts and formed a compact, most often unique, ball that was often located in the close proximity of the nucleus (Fig. 1B). As shown in Nomarski contrast, the vimentin bundles locally distended the cells, and were resistant to triton treatment that removes most of the proteins (Fig. 1C), and could thus be visualized by simple Coomassie blue staining (Fig. 1D).



View larger version (84K):
[in this window]
[in a new window]
  		Figure 1. (A–E) Immunofluorescence analysis of the vimentin organization in human primary fibroblasts derived from a healthy individual (A) and GAN patients (B, C, D, E). (A) The filamentous distribution of vimentin in the cytoplasm of control cells is similar in 10% serum and after 3 days in 0.1% of serum. (B) In 10% FCS, some GAN fibroblasts present partial aggregation of vimentin [the upper panel shows a comparable exposure time to (A) in order to reveal the presence of a well organized vimentin network in addition to the aggregates]. The vimentin bundles (arrowhead) are often single, close to the nucleus, and they locally distend the cells (as shown by Nomarski contrast, lower panel). The ovoid aggregates are resistant to triton treatment (C) and can therefore be visualized by Coomassie blue staining (D). (E) After 3 days in 0.1% FCS medium, new forms of vimentin aggregation appear (arrows): two bundles per cell, perinuclear rings and intermediate forms. The normal vimentin network, is still present but less visible (upper panel). (F–M) The destabilization of the vimentin network in GAN fibroblasts is enhanced in conditions that trigger the cells in the G0 phase, out of the cell cycle. WT fibroblasts (F, H, J, L) and GAN fibroblasts (G, I, K, M) are stained in red for vimentin and in green for the Ki-67 antigen, which is absent from cells that are in G0. The same vimentin disorganization (presence of two ovoid aggregates and perinuclear rings in most GAN fibroblasts) is observed after 3 days in 0.1% FCS (G) or after 3 days of confluence in 10% serum (I), and not in control fibroblasts (F and H, respectively). In these conditions, the Ki-67 antigen is not detected (G0). However, in 10% serum and when cells are not confluent (J, K), we found cells having vimentin aggregate that are negative (K1) and positive (K2) for the Ki-67 antigen, as are cells without aggregate (K3 and K4, respectively). On three independent counting, we found the same proportion of cells presenting vimentin aggregation in dividing and non-dividing cells (see text). Incubation during one day of confluence in 10% serum, when cells are still dividing, is not sufficient to enhance the destabilization of the vimentin network (L, M).

 
Nevertheless, in 10% FCS not all cytoplasmic vimentin was disorganized, since we found that the ovoid aggregates always coexisted with a well-formed network (Fig. 1B, upper panel). In addition, not all fibroblasts exhibited vimentin bundles. We found that the percentage of the cells presenting vimentin aggregation varied from one experiment to another, and from one patient to the other. With the exception of the fibroblasts of patient no. 4, which never presented aggregation as did the control fibroblasts, the proportion of cells with aggregates ranged from 3 to 15% (Fig. 2).



View larger version (36K):
[in this window]
[in a new window]
  		Figure 2. Percentage of primary fibroblasts presenting vimentin aggregation, in 10% serum and after 3 days in 0.1% of serum. The mutations and affected domain (BTB or kelch) are indicated below the patient numbers. A single mutation indicates that the patient is homozygous for this mutation. The question mark below patient no. 3 indicates that the second compound heterozygous mutation has not been identified. The percentage of positive cell has been calculated on three distinct experiments by counting twice 1000 cells in two different locations.

 
On low serum conditions, vimentin aggregates are enhanced and form perinuclear rings
After 3 days of culture in 0.1% FCS, we noticed both qualitative and quantitative changes in GAN fibroblasts. Different shapes of IF aggregation were observed, ranging from compact ovoid bundles, similar to those observed in 10% serum, except that there are two of them in most of the cells, usually with an opposite localization near the nucleus, to perinuclear rings. Intermediate forms are also present (Fig. 1E). The normal vimentin network in cells with aggregates is less visible in serum-deprived conditions (Fig. 1E, upper panel).

Here again, the fibroblasts of patient no. 4 never presented aggregation. In contrast, the other patients' fibroblasts revealed a 5–20-fold increase in the percentage of cells presenting vimentin aggregation, which ranged from 48 to 88% for cells in 0.1% FCS (Fig. 2). With the exception of patient no. 4, there was no correlation between the percentage of fibroblasts having aggregates both at 10 and 0.1% FCS and neither the type of mutation nor the domain affected in the GAN gene. The fibroblasts of patients no. 2 and 6, bearing homozygous missense and nonsense mutations in the Kelch domain, respectively, were used in the subsequent experiments and the fibroblasts of patient no. 4 (R15S missense mutation) were used only when indicated.

Prolonged confluence has the same effect on vimentin aggregation as low serum conditions
As serum starvation synchronizes the cells in the G0 phase of the cell cycle, we raised the possibility that the quantitative and qualitative effect on vimentin aggregation observed in 0.1% FCS could be explained by a recruitment of cells in G0. To address this question, we analyzed the Ki-67 antigen, which is expressed during all phases of the cell cycle and not in G0. As shown in Figure 1, all cells were indeed synchronized in G0 after 3 days in low serum condition (Fig. 1F and G). As prolonged confluence drives the cells into quiescence, we tested the possible effect of confluence on vimentin aggregation. We found that, as in low serum conditions, confluence for more than 3 days caused both a significant increase in the percentage of cells presenting aggregates and the appearance of perinuclear aggregates (Fig. 1H and I), while only one day of confluence had no effect (Fig. 1L and M). It is noticable that fibroblasts of patient no. 4 presented at 3 days of confluence rare cases of ring aggregates of vimentin, which were never observed in control fibroblasts.

However, when we tested for a correlation between the presence of vimentin aggregates and the cell cycle phase of non-confluent fibroblasts in 10% FCS, we found that all situations were possible, i.e. fibroblasts with aggregates that are positive (Fig. 1K2) or negative (Fig. 1K1) for the Ki-67 antigen and fibroblasts without aggregates that are also positive (Fig. 1K4) or negative (Fig. 1K3) for this antigen. A detailed counting of 100 cells in G0 and 100 cycling cells (positive for Ki-67) revealed no difference in the number of cells with vimentin aggregates. The results suggest that vimentin aggregation and disaggregation follow a kinetic that is slower than the cycling of fibroblasts.

The actin and microtubule networks are not altered in GAN fibroblasts
IFs disorganization can be induced by several drugs disrupting the MT network, such as colchicine, nocodazole and vinblastine (20). Thus, we wanted to determine if vimentin aggregation observed in GAN fibroblasts could be a consequence of MT destabilization. We analyzed the cytoskeletal components of GAN fibroblasts, cultivated both in 10 and 0.1% FCS. Both the MTs, the MT organizing centers (MTOCs), and the actin networks had normal distribution compared with the control cells (Fig. 3A and D), irrespective of the presence or not of aggregates and of the serum concentration (Fig. 3B, C, E and F). However, we observed that, in most cases, the vimentin aggregates were in the close vicinity of the MTOCs, both at 10 and 0.1% of serum (Fig. 3E and F), suggesting that an impaired interaction with the MT network drives the formation of the aggregates.



View larger version (94K):
[in this window]
[in a new window]
  		Figure 3. (A–F) The MF and the MT networks are not perturbed by vimentin aggregation, neither in 10% FCS nor in 0.1% FCS for 3 days. Co-staining of vimentin (mouse Vim 3B4 antibody in green) and actin (phalloidin TRITC) are seen in WT (A) and GAN (B, C) fibroblasts. (A) Ten percent FCS (a similar result is obtained in 0.1% FCS, not shown); (B) 10% FCS; (C) 0.1% FCS. A goat anti-human vimentin antibody (in green), that reveal some background staining, has been used to localize the aggregates, allowing us to look at the polymerized tubulin (revealed in red by mouse anti-{alpha}-tubulin B-5-1-2 antibody). As for the actin network, the MT network is correctly formed in GAN cells having or not vimentin aggregation, in 10% (E) or 0.1% (F) FCS, as in WT fibroblasts (D). We depicted the MT Organizing Centers (MTOCs) as white dots to show that the vimentin aggregates are usually located in their close vicinity. (G–J) MTs depolymerization induces the aggregation of all cytoplasmic vimentin in the vast majority of GAN cells. WT (G, I) and GAN (H, J) fibroblasts were stained for vimentin (upper panel) or for tubulin (lower panel). Overnight incubation with nocodazole (I, J) induces total depolymerization of the MT network, which subsequently induces retraction of the cells, and destabilizes the IFs network. While in control cells (I) vimentin forms ‘relaxed balls’ that fill all cytoplasmic areas (as revealed by the dashed lines representing the cellular outline, obtained from a merged image with Nomarski contrast); the very intense balls observed in GAN cells reflect a important retraction of all cytoplasmic IFs around the nucleus (J).

 
Vimentin aggregation is aggravated by MTs disruption
We analyzed the effect of MTs depolymerization on vimentin organization in GAN fibroblasts, by treating the cells with nocodazole. After overnight incubation in 5 µg/ml of nocodazole, all the MT network was depolymerized (Fig. 3I and J compared with G and H), and the elongated fibroblasts became retracted. As expected, the entire vimentin network was destabilized in control cells and in cells of patient no. 4, and appeared as relaxed coils (Fig. 3I). However, the nocodazole-induced disorganization of vimentin in the other GAN fibroblasts appeared very distinct. Indeed, the whole vimentin network became retracted around the nucleus in all cells and formed compact rings or very dense balls that were not solely accounted by the retraction of the cell shape but reflected a dynamic reorganization of the IFs (Fig. 3J). Although all MTs were depolymerized after 1 h of incubation with nocodazole, only few cells with disorganization of the vimentin network were observed at the same time (data not shown). In addition, the disorganization of the MT and IF networks was fully reversed after one day of incubation following the removal of nocodazole (not shown).

Subcellular localization of gigaxonin
We generated several specific antibodies against gigaxonin (Fig. 4) to examine its subcellular distribution. On immunoblots of Cos and HeLa cells transfected with a plasmid overexpressing gigaxonin, a band of the expected size (65 kDa) was detected with all polyclonal antibodies (Fig. 5, upper panel). Moreover, they were all able to immunoprecipitate the overexpressed gigaxonin revealed by the 1A2 monoclonal antibody (Fig. 5, lower panel). However, we failed to detect the endogenous gigaxonin from HeLa cells, human lymphoblastoid cell lines, primary fibroblasts, and mouse/rat neuronal cell lines. All bands that were detected with the crude antibodies and immuno-purified fractions were also found in total extracts of GAN fibroblasts and lymphoblastoïd cell types bearing homozygous nonsense mutations (patient nos 6 and 7, not shown). Similarly, immunostaining also failed to reveal the endogenous gigaxonin on HeLa cells, human primary fibroblasts, and human and mouse neuronal cell lines.



View larger version (12K):
[in this window]
[in a new window]
  		Figure 4. Schematic representation of gigaxonin and the immunogens used for antibody production. The BTB domain, at the amino part of gigaxonin and the six kelch repeats constituting the carboxyl kelch domain are represented. The N-terminal part of gigaxonin, comprising the BTB domain and the linker, has been produced in a prokaryotic expression system and was used to produce the polyclonal 1787 and the monoclonal 1A2 antibodies. The polyclonal antibodies 1781, 1777 and 1779 were raised against the peptides PF220, PF218 and PF219 (see Materials and Methods), respectively located in the BTB, the linker and the kelch domain.

 


View larger version (52K):
[in this window]
[in a new window]
  		Figure 5. Polyclonal antibodies directly detect and immunoprecipitate overexpressed gigaxonin. (Upper panel) Like the {alpha}-flag antibody, all our polyclonal antibodies detected a protein at the expected size in transfected Cos cells (5 µg of total protein extract). The specificity for gigaxonin is demonstrated by the absence of cross-reaction with preimmune serum (PI) or on non-transfected cells (Cos nt). Gigaxonin is less expressed in transfected HeLa cells (25 µg of protein extract). (Lower panel) Gigaxonin is immunoprecipitated by the 1781 polyclonal antibody, as well as by the {alpha}-flag antibody, on transfected HeLa cells (300 µg of protein extract), and revealed by the 1A2 monoclonal antibody. The controls c1, c2 and c3 correspond to the same experiment without antibody, using an anti-myotubularin polyclonal antibody for the immunoprecipitation, and without protein extract, respectively. Lanes 1 and 2 correspond to a direct detection on transfected Cos cells and HeLa cells, respectively.

 
When overexpressed, gigaxonin presented a fine or large granular distribution in the cytoplasm of HeLa cells (Fig. 6B), and Cos cells (not shown). The specificity of the detection was verified with flag-tagged gigaxonin revealed with anti-flag antibodies. When overexpressed, gigaxonin did not colocalize with either the IF, or MT or MF networks (Fig. 6).



View larger version (142K):
[in this window]
[in a new window]
  		Figure 6. Confocal microscopy of HeLa cells transfected with full flag tagged gigaxonin. Overexpressed gigaxonin forms fine and big dots in the cytoplasm of HeLa cells (B), and does not disturb nor colocalize with the MF, MT and IF networks (A), as shown in the merged pictures (C).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
When cultivated in normal serum condition (10% FCS), some fibroblasts of GAN patients display compact filamentous ovoid balls of vimentin. We confirmed that in most cases there is only one aggregate per cell, with a perinuclear localization (13,21) and that the aggregates co-exist with a well-organized vimentin network (13,22). It has been suggested that the bundles and normal vimentin network may be composed of different vimentin sub-population (13). We also confirmed that the fibroblasts with vimentin aggregate account for less than 20% of all cells (13,23), typically around 10% for our typical GAN cases. This is in contrast with other reports of 60% (14) to up to 90–95% (21) of GAN fibroblasts presenting aggregates. This difference is unlikely to be due to intra or inter-clonal variation as one study reported that there is limited clonal variation (23) and because the same fibroblasts (WG139) were described to present less than 20% of aggregates in one study (23) and about 90% in another study (22). Similarly, we found only 3.1% of aggregates in fibroblasts of patient no. 1, whereas the same primary line was reported to display 20% of aggregates in another study (13). Such discrepancies may result from different criteria to define aggregates, as we counted positive only completely formed ovoid aggregate. Nevertheless, the differences more likely reflect subtle variations of culture conditions, as they have been shown to greatly influence the percentage of aggregation. The low level of endogenous gigaxonin in fibroblasts, below the detection limit with our antibodies, precluded the study of the influence of residual mutant gigaxonin on the extent of vimentin aggregation. No noticeable difference was seen between the three cell lines bearing only truncating mutations (nonsense or splice site mutations) which are predicted to result in complete absence or severe alteration of the resulting protein.

We report here that fibroblasts of patients with typical GAN present a mean 7-fold increase (range 5–20-fold) of aggregates when grown during 3 days in 0.1% serum. The dependence on low serum condition was previously reported in one study that demonstrated that IF aggregation increased during time, and is inversely correlated to the percentage of serum in the medium (23), but was not reproduced in another study (14). Moreover, we found that low serum conditions resulted in the appearance of new types of aggregates. Indeed, in addition to the compact bundles which were duplicated in most cells, we observed compact perinuclear rings, with the presence of intermediate forms, that coexisted with a well-formed vimentin network.

As a prolonged culture in low serum triggers the cells in G0, we addressed whether vimentin aggregates may be preferentially formed during this phase, which would explain the low percentage of aggregates observed in normal serum conditions. In agreement with this hypothesis, we found that prolonged confluence, a condition that drives the cells into mitotic quiescence has the same effect on vimentin aggregation than prolonged culture in low serum. Nevertheless, we were not able to make any direct correlation between the presence of aggregates and the G0 phase in 10% FCS. These results are in agreement with the absence of correlation between DNA synthesis (as revealed by the mitotic index) and vimentin aggregation at different concentrations of serum (23).

We found that the actin network is not affected in GAN fibroblasts at any percentage of serum tested, in agreement with previous reports (13,21,22). A phenotype of the MT network would not be unexpected since its disruption with drugs such as colchicine, vinblastine, podophyllotoxin and nocodazole is known to induce IF aggregation in several cell types, such as fibroblasts, neurons and astrocytes. Contradictory studies reported that the MT network was normally distributed through the cytoplasm in GAN fibroblasts (22,23), or appeared disorganized and tangled in GAN cells, even in absence of vimentin bundles, with MTOCs not clearly visible (13). We found a MT network properly formed with clearly visible MTOCs in GAN fibroblasts cultivated both in 10 or 0.1% of serum, therefore excluding the possibility that vimentin aggregation is a consequence of the destabilization of the MT network. However, the striking positional association we found between the MTOCs and the vimentin aggregates at any percentage of serum is suggestive of a link between MTs and IFs that is mediated by gigaxonin. This hypothesis also is supported by the fact that, although destabilization of MTs induces the aggregation of whole cytoplasmic vimentin in control fibroblasts, we found that the absence of gigaxonin aggravates the phenotype. Indeed, upon treatment with nocodazole, all GAN cells exhibit densely packed balls and rings of vimentin around the nucleus instead of concentric relaxed filament for control cells. Even though the observation by Bousquet and colleagues of an abnormal MT organization in GAN cells was not confirmed by us and others, these authors also noted the vimentin phenotype aggravation upon nocodazole treatment (13), further supporting the link between gigaxonin and the MTs.

The implication of gigaxonin as a linker between the IF and MT networks has in part been elucidated by the recent identification of MAP1B as a neuronal partner of gigaxonin (24). MAP1B is a MT-associated protein expressed in developing neurons and maintains, when phosphorylated, the MTs of the growth cones in a dynamically unstable state necessary for axonogenesis (25). Cultured dorsal root ganglion neurons from MAP1B deficient mice present a reduction in axonal elongation, with an extension speed that is half reduced compared with control cells (26). Gigaxonin is able to directly bind through its kelch domain to the carboxyl end of the MAP1B light chain, and this association protects against drug-induced MT depolymerization (24). These authors found that, in neurons, endogenous gigaxonin colocalizes with MAP1B not only on cytoskeletal structures but also on spherical stuctures, suggesting that gigaxonin might be implicated in connecting vesicules with cytoskeleton. Only co-overexpression of gigaxonin and MAP1B allowed demonstrating the co-localization of the two proteins along MTs network in non-neuronal cells that lack endogenous MAP1B protein (24). Our finding with specific polyclonal antibodies that overexpressed gigaxonin in COS and Hela cells presents with a granular cytoplasmic distribution may suggest that, in the absence of MAP1B, gigaxonin is specifically associated with vesicular structures and that microtubule linking partners of gigaxonin in non neuronal cells are rate limiting.

The phenotypic presentation associated with specific missense mutations allows speculation on the relative importance of the domain they affect. The only two atypical families, in which the patients (four in each family) received an initial diagnosis of the milder disease Charcot–Marie–Tooth (1,2), bear a homozygous missense mutation, R15S and R138H, respectively (16). We have shown here that the R15S mutation is in addition associated with the absence of vimentin aggregation in fibroblasts. Our results indicate that the non-conserved arginine (R15) at the very beginning of the BTB domain, is not essential for gigaxonin function, but is undoubtedly necessary for its optimal functioning. The same conclusion probably also applies to the arginine (R138) located between the two major folds of the gigaxonin protein, the BTB and the Kelch domains. In contrast, we show that missense mutations in both the BTB domain (S79L) or the Kelch domain (E486K and R545C) are associated with vimentin aggregation and a typical early onset disease, indicating that the cellular and clinical phenotype related to gigaxonin dysfunction cannot be limited to one or the other domain, even though MAP1B interaction is mediated only by the Kelch domain (24). Although this interaction is the first evidence that gigaxonin participate to MT stability, the characteristic IF disorganization of GAN cells suggests that gigaxonin has additional interactors that would account for this phenotype. The search for new partners for both domains of gigaxonin should help elucidate its precise role in cytoskeleton networking.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Patients and cell lines
Patient nos 1–4 are from family nos 7, 8, 12 and 2 as reported in Bomont et al. (16), respectively; patient nos 5 and 6 are from family nos 2 and 1 as reported in Demir et al. (E. Demir et al., manuscript in preparation), and patient no. 7 is from family no. 16 as reported in Bomont et al. (27). All patients had a typical GAN presentation except patient no. 4, who belongs to a family in which a very mild form of GAN segregated (1). Primary fibroblasts were derived from skin biopsies taken from GAN patients and unrelated healthy individuals. Informed consent was obtained in all cases. The fibroblasts of patient no. 1–3 however previously described (13). Cells were propagated in DMEM with 10 or 0.1% FCS, supplemented with 40 µg/ml gentamycin and 1% penicillin/streptomycin (all Life Technologies, Grand Island, NY, USA). For MT-disrupting treatment, 5 µg/ml of nocodazole was added in 10% serum. HeLa cells were cultivated in Dulbeco 5% FCS with 40 µg/ml gentamycin and 1% penicillin/streptomycin.

Immunocytochemistry
After washing in PBS 1x, the cells were fixed for 10 min in 4% paraformaldehyde (PFA in PBS 1x), and washed again in PBS 1x. The cells were then incubated twice for 10 min in the permeabilization solution (PBS 0.1% Triton X-100), and washed in PBS 1x. Cells were incubated in the blocking solution (PBS 0.1% Triton 3% BSA) for 1 h, then washed in the permeabilization solution, and incubated with the primary antibody during 1 h in the blocking solution. After three washings with the permeabilization solution, the appropriate secondary antibody was applied for 1 h, followed by washing in PBS. The preparations were counterstained with Hoechst (Sigma, 5 µg/ml) for 10 s, washed twice with PBS and coverslips were mounted with an antifading solution (80% glycerol, 20% PBS 1x, 5% w/v propylgallate). All incubations were performed at room temperature.

Observations were carried out with a Leica microscope DMRA2 equipped with a video camera (Roper FX color) or with a confocal microscope (Leica TCS SP1). For visualization of the detergent resistant cytoskeleton, cells were incubated 30 min in 1% Triton X-100, 4% PEG 6000, 1 mM EDTA (in 0.1 M PIPES pH 6.8), washed in water, and fixed in 3.7% formaldehyde (in PBS). Cells were either permeabilized with cold methanol for 5 min for immunofluorescence, or stained for 5 min in coomassie solution (0.2% Coomassie blue R250, 7% acid acetic, 46.5% methanol) followed by destaining in 5% methanol, 5% acid acetic and air dried.

Antibodies against vimentin (mouse VIM 3B4 or goat anti-human vimentin), Ki-67 (rabbit) and {alpha}-tubulin (mouse B-5-1-2) were obtained from Novocastra (Newcastle, UK), Chemicon (Temecula, USA), DAKO (Glostrup, Danemark) and Sigma (Saint Louis, MO, USA), respectively. Phalloidin coupled to TRITC was obtained from Sigma. Goat anti-mouse Cy3, goat anti-mouse Alexa 488 and donkey anti-mouse Cy3 secondary antibodies for immunofluorescence were obtained from Jackson Immunoresearch (West Grove, USA) and donkey anti-goat alexa 488 antibodies were obtained from Molecular Probes (Leiden, The Netherlands).

Quantification of vimentin aggregation
The percentage of cells presenting vimentin aggregation was scored in a total of 1000 cells, as determined by Hoechst staining. Fibroblasts were scored positive only when they contained ovoid aggregates completely formed, perinuclear rings or intermediate forms that could be unambiguously detected. Two counts were performed on three independent experiments each.

Production of antibodies against gigaxonin
N-terminal part of gigaxonin (nt 1–670) was amplified by PCR with primers containing the BamHI restriction site in forward primer and NotI-BamHI sites in the reverse primer, and cloned in the procaryotic expressing vector pGEX4t3, after BamHI digestion. After induction of expression by IPTG, the GST fused proteins were purified on Gluthathione Sepharose 4B (Pharmacia Biotech). The recombinant protein was injected both in mice and rabbits. In addition, three peptides were selected according to their antigenicity index (Jameson-Wolf) and synthetized: PF220 in the BTB domain (RTKLNYNPPKDDGSTYK), PF218 in the linker domain (TEYLETHFRDVSSTEEFLE) and PF219 in the Kelch domain (DENKQTLSSGEKYDPDANT). A cysteine residue was included at the amino end of each peptide for cross-linking to the carrier ovalbumin (Pierce, Imject Maleimide Activated Ovalbumin), prior to injection into rabbits. All polyclonal antibodies were purified with the corresponding immobilized antigen (Pierce, SulfoLink Coupling Gel). Mouse hybridomas directed against the N-terminal part of gigaxonin were constructed and selected.

Transfection
The full gigaxonin sequence (nt 1–1794), was amplified by PCR and cloned in the eucaryotic expressing vector pTL10Sflag (derived from pSG5 to include a flag epitope; gift from D. Devys); 0.2 µg/cm2 of plasmid were incubated with 0.125 mM CaCl2, 1x BBS for 20 min at room temperature, and then added to cultured cells. On the following day, the precipitates were washed and immunocytochemistry or proteins extractions were performed the day after.

Immunoprecipitation and western blot
Cells were lysed by sonication in 50 mM Tris–HCl, 150 mM NaCl, 1% NP40 with a cocktail of protease inhibitors (PIC 1x). For immunoprecipitation experiments, 300 µg of transfected HeLa cells were mixed overnight at 4°C under mild agitation with 5 µl of rabbit serum. After 1 h incubation with 40 µl of Protein A Sepharose (Pharmacia Biotech), the immunocomplexes were collected by centrifugation, washed four times in 50 mM Tris–HCl, 300 mM NaCl, 1% NP40, PIC 1x during 15 min, and resuspended in 20 µl of loading buffer (8% SDS, 40% glycerol, 240 mM Tris, 0.004% bromophenol blue).

Proteins were separated on 8% SDS–polyacrylamide gel and electrophoretically transferred onto nitrocellulose membranes. The membranes were blocked with 5% non-fat milk in PBS, 0.05% Tween20 for 30 min, then incubated with the primary antibody in 0.5% milk ({alpha}-flag 1:5000, 1A2 ascite and polyclonal antibodies against gigaxonin 1:1000) at 4°C overnight. After washings in PBS 0.05% Tween20, the secondary antibody (goat anti-mouse or anti-rabbit antibody coupled to peroxidase, Jackson Immunoresearch; 1:10 000) was incubated for 1 h at room temperature, and immunoreactivity was revealed using the Super Signal chemoluminescence kit (Pierce).


   ACKNOWLEDGEMENTS
 
We wish to thank Professor J.-L. Mandel for support, and Drs M.-M. Portier, E. Demir, M. Tazir, S. Belal and F. Hentati for the fibroblast cell lines. We also wish to thank all the members of the Imaging Center, the cell culture, and antibody facilities at the IGBMC for their kind and very helpful support, and Y. Trottier for helpful discussions. This work was supported by funds from the Centre National de la Recherche Scientifique (CNRS), the Institut National de la Santé et de la Recherche Médicale (INSERM), the Collège de France, the Hôpital Universitaire de Strasbourg (HUS) and the Association Française contre les Myopathies (AFM). P.B. was a recipient of a fellowship from AFM.


   FOOTNOTES
 
* To whom correspondence should be addressed. Tel: +33 388653399; Fax: +33 388653246; Email: mkoenig{at}igbmc.u-strasbg.fr Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Ben Hamida, M., Hentati, F. and Ben Hamida, C. (1990) Giant axonal neuropathy with inherited multisystem degeneration in a Tunisian kindred. Neurology, 40, 245–250.[Abstract/Free Full Text]

  2. Zemmouri, R., Azzedine, H., Assami, S., Kitouni, N., Vallat, J.M., Maisonobe, T., Hamadouche, T., Kessaci, M., Mansouri, B., Le Guern, E. et al. (2000) Charcot–Marie–Tooth 2-like presentation of an Algerian family with giant axonal neuropathy. Neuromusc. Disord., 10, 592–598.[CrossRef][Web of Science][Medline]

  3. Fois, A., Balestri, P., Farnetani, M.A., Berardi, R., Mattei, R., Laurenzi, E., Alessandrini, C., Gerli, R., Ribuffo, A. and Calvieri, S. (1985) Giant axonal neuropathy. Endocrinological and histological studies. Eur. J. Pediatr., 144, 274–280.[CrossRef][Web of Science][Medline]

  4. Koch, T., Schultz, P., Williams, R. and Lampert, P. (1977) Giant axonal neuropathy: a childhood disorder of microfilaments. Ann. Neurol., 1, 438–451.[CrossRef][Web of Science][Medline]

  5. Mohri, I., Taniike, M., Yoshikawa, H., Higashiyama, M., Itami, S. and Okada, S. (1998) A case of giant axonal neuropathy showing focal aggregation and hypophosphorylation of intermediate filaments. Brain Dev., 20, 594–597.[CrossRef][Web of Science][Medline]

  6. Takebe, Y., Koide, N. and Takahashi, G. (1981) Giant axonal neuropathy: report of two siblings with endocrinological and histological studies. Neuropediatrics, 12, 392–404.[Web of Science][Medline]

  7. Treiber-Held, S., Budjarjo-Welim, H., Reimann, D., Richter, J., Kretzschmar, H.A. and Hanefeld, F. (1994) Giant axonal neuropathy: a generalized disorder of intermediate filaments with longitudinal grooves in the hair. Neuropediatrics, 25, 89–93.[Web of Science][Medline]

  8. Herrmann, H. and Aebi, U. (2000) Intermediate filaments and their associates: multi-talented structural elements specifying cytoarchitecture and cytodynamics. Curr. Opin. Cell Biol., 12, 79–90.[CrossRef][Web of Science][Medline]

  9. Donaghy, M., King, R.H., Thomas, P.K. and Workman, J.M. (1988) Abnormalities of the axonal cytoskeleton in giant axonal neuropathy. J. Neurocytol., 17, 197–208.[CrossRef][Web of Science][Medline]

  10. Guazzi, G.C., Malandrini, A., Gerli, R. and Federico, A. (1991) Giant axonal neuropathy in 2 siblings: a generalized disorder of intermediate filaments. Eur. Neurol., 31, 50–56.[Web of Science][Medline]

  11. Kretzschmar, H.A., Berg, B.O. and Davis, R.L. (1987) Giant axonal neuropathy. A neuropathological study. Acta Neuropathol., 73, 138–144.[CrossRef][Medline]

  12. Peiffer, J., Schlote, W., Bischoff, A., Boltshauser, E. and Muller, G. (1977) Generalized giant axonal neuropathy: a filament-forming disease of neuronal, endothelial, glial, and schwann cells in a patient without kinky hair. Acta Neuropathol. (Berl.), 40, 213–218.[CrossRef][Medline]

  13. Bousquet, O., Basseville, M., Vila-Porcile, E., Billette de Villemeur, T., Hauw, J.J., Landrieu, P. and Portier, M.M. (1996) Aggregation of a subpopulation of vimentin filaments in cultured human skin fibroblasts derived from patients with giant axonal neuropathy. Cell Motil. Cytoskeleton, 33(2), 115–129.

  14. Manetti, R., Ceccarini, C., Guazzi, G., Federico, A., Tiezzi, A., Bugnoli, M. and Ceccarini, E.C. (1987) Characterization of the intermediate filament apparatus in skin fibroblasts from patients with giant axonal neuropathy: effect of trypsin. Cell Motil. Cytoskeleton, 8, 55–60.[CrossRef][Web of Science][Medline]

  15. Pena, S.D. (1982) Giant axonal neuropathy: an inborn error of organization of intermediate filaments. Muscle Nerve, 5, 166–172.[CrossRef][Web of Science][Medline]

  16. Bomont, P., Cavalier, L., Blondeau, F., Ben Hamida, C., Belal, S., Tazir, M., Demir, E., Topaloglu, H., Korinthenberg, R., Tuysuz, B. et al. (2000) The gene encoding gigaxonin, a new member of the cytoskeletal BTB/kelch repeat family, is mutated in giant axonal neuropathy. Nat. Genet., 26, 370–374.[CrossRef][Web of Science][Medline]

  17. Adams, J., Kelso, R. and Cooley, L. (2000) The kelch repeat superfamily of proteins: propellers of cell function. Trends Cell Biol., 10, 17–24.[CrossRef][Web of Science][Medline]

  18. Bardwell, V.J. and Treisman, R. (1994) The POZ domain: a conserved protein–protein interaction motif. Genes Dev., 8, 1664–1677.[Abstract/Free Full Text]

  19. Zollman, S., Godt, D., Prive, G.G., Couderc, J.L. and Laski FA. (1994) The BTB domain, found primarily in zinc finger proteins, defines an evolutionarily conserved family that includes several developmentally regulated genes in Drosophila. Proc. Natl Acad. Sci. USA, 91, 10717–10721.[Abstract/Free Full Text]

  20. Lazarides, E. (1980) Intermediate filaments as mechanical integrators of cellular space. Nature, 283, 249–256.[CrossRef][Medline]

  21. Pena, S.D. (1981) Giant axonal neuropathy: intermediate filament aggregates in cultured skin fibroblasts. Neurology, 31, 1470–1473.[Abstract/Free Full Text]

  22. Pena, S.D., Opas, M., Turksen, K., Kalnins, V.I. and Carpenter, S. (1983) Immunocytochemical studies of intermediate filament aggregates and their relationship to microtubules in cultured skin fibroblasts from patients with giant axonal neuropathy. Eur. J. Cell Biol., 31, 227–234.[Web of Science][Medline]

  23. Klymkowsky, M.W. and Plummer, D.J. (1985) Giant axonal neuropathy: a conditional mutation affecting cytoskeletal organization. J. Cell Biol., 100, 245–250.[Abstract/Free Full Text]

  24. Ding, J., Liu, J.J., Kowal, A.S., Nardine, T., Bhattacharya, P., Lee, A. and Yang, Y. (2002) Microtubule-associated protein 1B: a neuronal binding partner for gigaxonin. J. Cell Biol., 158, 427–433.[Abstract/Free Full Text]

  25. Goold, R.G., Owen, R. and Gordon-Weeks, P.R. (1999) Glycogen synthase kinase 3beta phosphorylation of microtubule-associated protein 1B regulates the stability of microtubules in growth cones. J. Cell Sci., 112, 3373–3384.[Abstract]

  26. Gonzalez-Billault, C., Owen, R., Gordon-Weeks, P.R. and Avila, J. (2002) Microtubule-associated protein 1B is involved in the initial stages of axonogenesis in peripheral nervous system cultured neurons. Brain Res., 943, 56–67.[CrossRef][Web of Science][Medline]

  27. Bomont, P., Ioos, C., Yalcinkaya, C., Korinthenberg, R., Vallat, J.M., Assami, S., Munnich, A., Chabrol, B., Kurlemann, G., Tazir, M. and Koenig, M. (2003) Identification of Seven Novel Mutations in the GAN gene. Hum. Mutat. Mutation in Brief #593 online.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Hum Mol GenetHome page
D. W. Cleveland, K. Yamanaka, and P. Bomont
Gigaxonin controls vimentin organization through a tubulin chaperone-independent pathway
Hum. Mol. Genet., April 15, 2009; 18(8): 1384 - 1394.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
B. Cui, C. Wu, L. Chen, A. Ramirez, E. L. Bearer, W.-P. Li, W. C. Mobley, and S. Chu
One at a time, live tracking of NGF axonal transport using quantum dots
PNAS, August 21, 2007; 104(34): 13666 - 13671.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
D. D. Zhang, S.-C. Lo, Z. Sun, G. M. Habib, M. W. Lieberman, and M. Hannink
Ubiquitination of Keap1, a BTB-Kelch Substrate Adaptor Protein for Cul3, Targets Keap1 for Degradation by a Proteasome-independent Pathway
J. Biol. Chem., August 26, 2005; 280(34): 30091 - 30099.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (17)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Bomont, P.
Right arrow Articles by Koenig, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bomont, P.
Right arrow Articles by Koenig, M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?